ci: compute commit range for rerun workflows (#27775)

On push builds, CircleCI provides `CIRCLE_COMPARE_URL`, which we use to
extract the commit range for a given build. When a workflow is rerun
(e.g. to recover from a flaked job), `CIRCLE_COMPARE_URL` is not
defined, causing some jobs to fail.

This commit fixes it by retrieving the compare URL from the original
workflow. It uses a slow process involving a (potentially large) number
of requests to CircleCI API.
It depends on the (undocumented) fact, that the `workspace_id` is the
same on all rerun workflows and the same as the original `workflow_id`.

PR Close #27775

.circleci/env.sh

@@ -1,7 +1,10 @@
 #!/usr/bin/env bash
 
-# Load helpers and make them available everywhere (through `$BASH_ENV`).
+# Variables
 readonly envHelpersPath="`dirname $0`/env-helpers.inc.sh";
+readonly getCommitRangePath="`dirname $0`/get-commit-range.js";
+
+# Load helpers and make them available everywhere (through `$BASH_ENV`).
 source $envHelpersPath;
 echo "source $envHelpersPath;" >> $BASH_ENV;
 
@@ -16,9 +19,10 @@ setPublicVar CI_AIO_MIN_PWA_SCORE "95";
 # This is the branch being built; e.g. `pull/12345` for PR builds.
 setPublicVar CI_BRANCH "$CIRCLE_BRANCH";
 setPublicVar CI_COMMIT "$CIRCLE_SHA1";
-# `CI_COMMIT_RANGE` will only be available when `CIRCLE_COMPARE_URL` is also available,
+# `CI_COMMIT_RANGE` will only be available when `CIRCLE_COMPARE_URL` is also available (or can be
-# i.e. on push builds (a.k.a. non-PR builds). That is fine, since we only need it in push builds.
+# retrieved via `get-compare-url.js`), i.e. on push builds (a.k.a. non-PR, non-scheduled builds and
-setPublicVar CI_COMMIT_RANGE "$(sed -r 's|^.*/([0-9a-f]+\.\.\.[0-9a-f]+)$|\1|i' <<< ${CIRCLE_COMPARE_URL:-})";
+# rerun workflows of such builds). That is fine, since we only need it in push builds.
+setPublicVar CI_COMMIT_RANGE "`[[ ${CIRCLE_PR_NUMBER:-false} != false ]] && echo "" || node $getCommitRangePath "$CIRCLE_BUILD_NUM" "$CIRCLE_COMPARE_URL"`";
 setPublicVar CI_PULL_REQUEST "${CIRCLE_PR_NUMBER:-false}";
 setPublicVar CI_REPO_NAME "$CIRCLE_PROJECT_REPONAME";
 setPublicVar CI_REPO_OWNER "$CIRCLE_PROJECT_USERNAME";

.circleci/get-commit-range.js

@@ -0,0 +1,149 @@
+#!/usr/bin/env node
+
+/**
+ * **Usage:**
+ * ```
+ * node get-commit-range <build-number> [<compare-url> [<circle-token>]]
+ * ```
+ *
+ * Returns the value of the `CIRCLE_COMPARE_URL` environment variable (if defined) or, if this is
+ * not a PR build (i.e. `CIRCLE_PR_NUMBER` is not defined), retrieves the equivalent of
+ * `CIRCLE_COMPARE_URL` for jobs that are part of a rerun workflow.
+ *
+ * **Context:**
+ * CircleCI sets the `CIRCLE_COMPARE_URL` environment variable (from which we can extract the commit
+ * range) on push builds (a.k.a. non-PR, non-scheduled builds). Yet, when a workflow is rerun
+ * (either from the beginning or from failed jobs) - e.g. when a job flakes - CircleCI does not set
+ * the `CIRCLE_COMPARE_URL`.
+ *
+ * **Implementation details:**
+ * This script relies on the fact that all rerun workflows share the same CircleCI workspace and the
+ * (undocumented) fact that the workspace ID happens to be the same as the workflow ID that first
+ * created it.
+ *
+ * For example, for a job on push build workflow, the CircleCI API will return data that look like:
+ * ```js
+ * {
+ *   compare: 'THE_COMPARE_URL_WE_ARE_LOOKING_FOR',
+ *   //...
+ *   previous: {
+ *     // ...
+ *     build_num: 12345,
+ *   },
+ *   //...
+ *   workflows: {
+ *     //...
+ *     workflow_id: 'SOME_ID_A',
+ *     workspace_id: 'SOME_ID_A',  // Same as `workflow_id`.
+ *   }
+ * }
+ * ```
+ *
+ * If the workflow is rerun, the data for jobs on the new workflow will look like:
+ * ```js
+ * {
+ *   compare: null,  // ¯\_(ツ)_/¯
+ *   //...
+ *   previous: {
+ *     // ...
+ *     build_num: 23456,
+ *   },
+ *   //...
+ *   workflows: {
+ *     //...
+ *     workflow_id: 'SOME_ID_B',
+ *     workspace_id: 'SOME_ID_A',  // Different from current `workflow_id`.
+ *                                 // Same as original `workflow_id`. \o/
+ *   }
+ * }
+ * ```
+ *
+ * This script uses the `previous.build_num` (which points to the previous build number on the same
+ * branch) to traverse the jobs backwards, until it finds a job from the original workflow. Such a
+ * job (if found) should also contain the compare URL.
+ *
+ * **NOTE 1:**
+ * This is only useful on workflows which are created by rerunning a workflow for which
+ * `CIRCLE_COMPARE_URL` was defined.
+ *
+ * **NOTE 2:**
+ * The `circleToken` will be used for CircleCI API requests if provided, but it is not needed for
+ * accessing the read-only endpoints that we need (as long as the current project is FOSS and the
+ * corresponding setting is turned on in "Advanced Settings" in the project dashboard).
+ *
+ * ---
+ * Inspired by https://circleci.com/orbs/registry/orb/iynere/compare-url
+ * (source code: https://github.com/iynere/compare-url-orb).
+ */
+
+// Imports
+const {get: httpsGet} = require('https');
+
+// Constants
+const API_URL_BASE = 'https://circleci.com/api/v1.1/project/github/angular/angular';
+const COMPARE_URL_RE = /^.*\/([0-9a-f]+\.\.\.[0-9a-f]+)$/i;
+
+// Run
+_main(process.argv.slice(2));
+
+// Helpers
+async function _main([buildNumber, compareUrl = '', circleToken = '']) {
+  try {
+    if (!buildNumber || isNaN(buildNumber)) {
+      throw new Error(
+          'Missing or invalid arguments.\n' +
+          'Expected: buildNumber (number), compareUrl? (string), circleToken? (string)');
+    }
+
+    if (!compareUrl) {
+      compareUrl = await getCompareUrl(buildNumber, circleToken);
+    }
+
+    const commitRangeMatch = COMPARE_URL_RE.exec(compareUrl)
+    const commitRange = commitRangeMatch ? commitRangeMatch[1] : '';
+
+    console.log(commitRange);
+  } catch (err) {
+    console.error(err);
+    process.exit(1);
+  }
+}
+
+function getBuildInfo(buildNumber, circleToken) {
+  console.error(`BUILD ${buildNumber}`);
+  const url = `${API_URL_BASE}/${buildNumber}?circle-token=${circleToken}`;
+  return getJson(url);
+}
+
+async function getCompareUrl(buildNumber, circleToken) {
+  let info = await getBuildInfo(buildNumber, circleToken);
+  const targetWorkflowId = info.workflows.workspace_id;
+
+  while (info.workflows.workflow_id !== targetWorkflowId) {
+    info = await getBuildInfo(info.previous.build_num, circleToken);
+  }
+
+  return info.compare || '';
+}
+
+function getJson(url) {
+  return new Promise((resolve, reject) => {
+    const opts = {headers: {Accept: 'application/json'}};
+    const onResponse = res => {
+      const statusCode = res.statusCode || -1;
+      const isSuccess = (200 <= statusCode) && (statusCode <= 400);
+      let responseText = '';
+
+      res.
+        on('error', reject).
+        on('data', d => responseText += d).
+        on('end', () => isSuccess ?
+          resolve(JSON.parse(responseText)) :
+          reject(`Error getting '${url}' (status ${statusCode}):\n${responseText}`));
+    };
+
+    httpsGet(url, opts, onResponse).
+      on('error', reject).
+      end();
+  });
+}